TIDAL Can Take Your Royalties Away — Even Though Its Own AI Detector Admits It Can Be Wrong

There is a sentence in TIDAL’s current AI policy that should make every independent artist stop and read the small print: recordings that TIDAL identifies as wholly AI-generated are not eligible for royalty attribution. TIDAL presents this as a way of ensuring that royalties reach original works directly produced, written and performed by people, and there is an entirely legitimate problem behind that objective because industrial-scale AI content, impersonation and artificial streaming can divert money and attention away from genuine creators.

But the important issue is what happens when TIDAL’s identification is wrong. Its own terms state that AI identification and labelling are performed on a best-efforts basis, that TIDAL does not warrant the accuracy or completeness of the label, and that AI detection technology can produce false positives or false negatives. That combination should fundamentally change how artists understand the system. TIDAL is not saying, “We have definitively established that this recording is AI-generated.” It is effectively saying, “We have a detection system that we believe is sufficiently useful to classify recordings, but we do not guarantee that the classification is correct.” Yet the economic consequence of that classification can fall directly on the artist.

TIDAL has spent years cultivating an identity built around artist support, higher-fidelity listening and a closer relationship between musicians and fans, but corporate branding is not the same thing as structural independence. The moment a platform can decide that a recording belongs outside its royalty system, the artist is no longer dealing with a neutral distribution pipe.

They are dealing with an institution that has created its own rules, built its own detection machinery and reserved the authority to enforce those rules. TIDAL’s terms explicitly reserve the right to exclude content it determines to be AI-generated from royalty calculations, while its content guidelines state that AI-related violations may be labelled, demonetised or removed at TIDAL’s sole discretion and without prior notice. That is a remarkable concentration of power over creative work and creative income.

The company gets to establish the category, operate the detection process, decide what constitutes a violation and determine the commercial consequences. The artist gets an appeal route after the decision has already been made.

The phrase “best efforts” sounds harmless until money is attached to it. TIDAL says that its AI identification and labelling operate on a best-efforts basis and that its technology can generate false positives and false negatives. That means the company knows its system is not infallible. Every automated classification system has limitations, of course, but the consequences here are unusually important because the output is connected to a commercial decision.

A wrong recommendation might annoy an artist. A wrong AI classification can potentially alter how listeners perceive the recording and, under TIDAL’s policy, whether the recording receives royalty attribution at all. The distinction matters because TIDAL is not merely identifying a technical characteristic in the audio. It is potentially making a financial determination based on that identification.

The disturbing part is not that TIDAL wants to stop fraudulent AI music from earning money. That is a defensible objective. The disturbing part is the architecture surrounding the decision. TIDAL’s terms give it discretion to exclude content it determines to be AI-generated from royalty calculations.

The platform therefore has a mechanism through which an automated technological assessment can become an economic decision. The artist does not get to participate in the initial classification. There is no requirement in the published policy that TIDAL publicly disclose a detailed evidentiary record for every classification. Instead, the artist is directed toward support if they believe the track has been identified incorrectly.

That is backwards from the perspective of accountability. When the platform controls the evidence, the algorithm and the money, the creator should not have to begin by assuming that the platform is right and then prove that it is wrong.

TIDAL’s admission that false positives can occur is particularly important because independent music production has become extraordinarily complicated. Modern producers use samplers, synthesisers, machine-learning tools, AI-assisted sound design, stem separation, restoration technology, mastering software, generative textures, virtual instruments, procedural systems and countless combinations of digital tools.

A recording can contain material generated or transformed by software without the recording itself being wholly generated by AI. TIDAL’s published definition of AI-generated content focuses on audio that is wholly or substantially generated by generative artificial intelligence with limited or no direct human creative input beyond an initial prompt or similar instruction. That is a much more nuanced concept than simply detecting something that “sounds like AI.” The danger appears when technological detection becomes a substitute for understanding the production process.

There is an enormous difference between music made by AI and music made by a human using AI as one tool among many. TIDAL’s policy recognises this distinction in its definition, but an automated detector cannot necessarily reconstruct the entire history of a recording simply by analysing the final audio.

A detector may identify characteristics associated with generated content, but it does not automatically know who wrote the underlying material, who arranged the composition, who edited the recording, who made the production decisions, who transformed the generated material, how many human interventions occurred or whether AI was merely one component inside a much larger creative process.

That is why a platform that wants to connect detection with financial consequences needs a substantially higher standard than “our system detected it.” TIDAL’s own terms acknowledge that the system can be wrong. The platform therefore needs to explain why its classification should carry financial authority when the technology itself is not guaranteed to be accurate.

The UNIDARK catalogue is a useful example of the complexity TIDAL’s system has to confront. UNIDARK openly uses AI textures alongside a multitude of other production tools, but the project is not simply a catalogue of songs generated wholesale by a machine.

The music is developed through human creative direction, composition, arrangement, production, editing and extensive decision-making, and the catalogue is PRS registered according to the project’s records. Several UNIDARK tracks have nevertheless received AI labels on TIDAL. That does not by itself establish that TIDAL violated its policy, because the platform’s determination concerns its own classification criteria and the specific characteristics it detected.

But it raises a serious question that TIDAL needs to answer: what exactly distinguishes a wholly AI-generated recording from a human-created recording that incorporates AI-assisted textures and tools? If the answer is hidden inside a proprietary detection process, then the artist is being asked to accept a potentially consequential classification without being given enough information to independently evaluate how the conclusion was reached.

It would be inaccurate to suggest that PRS registration itself proves that a recording cannot involve AI. Copyright, authorship, performance, production and platform classification are different questions.

What registration does demonstrate is that the work exists within a formal rights and authorship framework rather than simply appearing anonymously inside a platform’s database. That context matters when a streaming service applies an automated label that can influence monetisation. The artist is not merely an audio file uploaded into a machine. There is a creative and rights history behind the recording. A serious moderation process should be capable of considering that history rather than treating the audio signal as the entire truth.

This is the contradiction that should anger independent musicians most. TIDAL wants artists and listeners to treat its classifications as meaningful enough to affect the treatment of recordings, but its own legal language protects the company from guaranteeing that the classifications are accurate. This creates an asymmetry of certainty.

TIDAL can act with sufficient certainty to classify the music, while simultaneously retaining sufficient uncertainty to disclaim responsibility if the classification is inaccurate. The artist experiences the decision as concrete. The corporation describes the underlying technology as fallible. That is an extremely comfortable position for the platform and a potentially uncomfortable one for the creator.

The entire debate becomes more serious once royalties enter the equation. TIDAL explicitly states that recordings it identifies as wholly AI-generated are not eligible for royalty attribution. This is not a philosophical label. It can have a direct economic consequence.

The artist has created something that generated listening activity, but the platform can decide that the recording does not qualify for royalty attribution because of its determination about how that recording was created. If that determination is accurate and the recording really is wholly AI-generated, TIDAL can defend the policy as a mechanism for protecting the royalty pool. But if the determination is wrong, the same mechanism becomes a barrier between the artist and their money.

TIDAL does provide a route for artists who believe their music has been incorrectly identified. Its support documentation instructs artists to contact support when a track has been erroneously tagged, and its moderation guidelines say appeals can trigger human review. That is better than having no appeal mechanism at all. But an appeal does not erase the fundamental imbalance. The platform’s automated system gets the first move. The artist gets the burden of noticing the problem and challenging it. The classification may already have affected the recording’s treatment before the appeal is completed.

A genuinely creator-centred system would make the correction process fast, transparent and financially restorative when the platform is wrong. Otherwise, “appeal” risks becoming corporate language for “convince us that our system made a mistake.”

This is one of the questions artists should be asking TIDAL directly. Suppose a recording is wrongly classified as wholly AI-generated and therefore excluded from royalty attribution. Suppose the artist appeals. Suppose the platform eventually agrees that the classification was wrong. What happens to the royalties associated with the affected streams? TIDAL’s public AI documentation tells artists how to request a review, but the published material does not provide a detailed public accounting framework explaining precisely how disputed royalty amounts are reconstructed and repaid after an erroneous AI classification. That gap deserves scrutiny. A moderation system connected to money needs a financial correction mechanism just as much as it needs a content appeal mechanism.

Artists have spent years fighting over streaming economics because the basic relationship between listening activity and creator income is difficult to understand from outside the platform-distributor-label chain. TIDAL itself states that royalty reporting is provided to distribution partners and that the distributor or label is generally the artist’s point of contact for royalty data and disbursement. That already creates layers between the artist and the platform. Now add another layer of automated classification.

The artist may not even see the underlying platform decision immediately, while the financial data travels through distributors and rights holders. The result is a system where the creator can be several steps removed from the original decision that affects the money.

The AI policy is not operating in isolation. TIDAL also states that royalties are not paid for tracks identified as having artificial streams and that those tracks can be excluded from royalty statements for applicable accounting periods. Again, the legitimate goal is understandable: nobody should profit from fraudulent streaming. But the broader lesson for artists is that streaming platforms increasingly operate as financial gatekeepers using automated or platform-level detection systems.

AI classification is simply the latest version of a much larger trend. The platform decides what counts as legitimate activity, what counts as suspicious activity and what activity receives economic recognition. That is a tremendous amount of power concentrated inside a company whose primary relationship with the artist is contractual and technological rather than personal.

The wording of TIDAL’s content guidelines should be read carefully. The company says AI-related content that violates its guidelines may be labelled, demonetised or removed at its sole discretion and without prior notice. It also says automated systems scan and filter content upon upload to identify potential violations, with human review available for flagged content and appeals.

This means TIDAL has built a moderation structure where automation is embedded directly into the pipeline. The platform is not waiting for an artist dispute before evaluating the content. It is actively scanning uploads and making classifications. That may be operationally necessary at scale, but operational necessity does not make the underlying power harmless.

Automation is attractive to a company because it scales far beyond what a human moderation workforce could achieve manually. TIDAL can scan enormous quantities of content without assigning an individual employee to every recording. That is economically efficient.

But the cost of a false positive does not fall equally on both sides. For TIDAL, an erroneous classification can be one more moderation event in an enormous database. For an independent musician, it can involve lost royalties, lost visibility, damaged credibility and hours spent trying to obtain an explanation. The corporation receives the efficiency of automation. The individual creator can receive the consequences of its mistakes. That is the economic imbalance hiding underneath the technical language.

TIDAL describes its AI policy as evolving alongside the technology. It currently focuses action on content it determines to be wholly AI-generated because of limitations in AI detection and the need to avoid false positives, while saying it may expand its actions to substantially AI-generated content as detection becomes more reliable. That should concern every artist using technological tools in their production. The rules are not static.

The boundary can move. What is acceptable today may receive a different treatment tomorrow. An artist who builds a production workflow around current platform assumptions may discover that the platform has changed the definition after the music has already been released.

If TIDAL eventually expands enforcement beyond wholly AI-generated music, the word “substantially” becomes enormously important. How much AI involvement qualifies? What happens when a human writes a composition but uses generated textures? What happens when a producer generates dozens of possibilities and manually selects and transforms one? What happens when AI is used for a vocal layer, sound design, restoration or experimental texture? What happens when a human performance is processed through machine-learning software? These are not hypothetical edge cases anymore.

Modern music production is increasingly hybrid. TIDAL’s current policy deliberately avoids broader enforcement while its detection technology remains limited, but the company explicitly leaves open the possibility of expanding its actions later. That means today’s uncertainty could become tomorrow’s monetisation dispute.

AI is merely the latest justification for expanding the platform’s ability to inspect and classify creative work. The deeper problem is the assumption that the corporation should have the final word over what a recording is and whether that recording deserves economic recognition. TIDAL can legitimately fight fraud. It can legitimately refuse to reward impersonation. It can legitimately protect its royalty pool. But those powers need boundaries. A corporation should not be able to transform a probabilistic technological assessment into a financial consequence while simultaneously disclaiming responsibility for the accuracy of the underlying technology.

That is where the rot begins.

TIDAL did not write the song. TIDAL did not compose the arrangement. TIDAL did not operate the instruments. TIDAL did not make the production decisions. TIDAL did not spend the hours building the recording. Yet the platform can decide whether it believes that recording qualifies for royalty attribution. This is one of the strangest developments in modern music: companies that did not create the work increasingly control the conditions under which the work is recognised economically. Streaming platforms call this moderation. From the artist’s perspective, it can look like a new form of gatekeeping.

Listeners usually see a song title, artwork, artist name and play button. They do not see the automated systems scanning the content. They do not see the platform’s internal classifications. They do not see the contractual language reserving discretion. They do not see the distributor reporting chain. They do not see the royalty accounting process. They simply press play. That simplicity is the entire appeal of streaming, but it also hides enormous institutional power. The platform appears frictionless precisely because most of the complexity has been pushed behind the interface.

The artist lives on the other side of that interface.

The artist experiences the consequences.

A stream is not ownership. A profile is not ownership. A playlist placement is not ownership. A verified artist page is not ownership. A presence on TIDAL is not independence. It is access to someone else’s infrastructure. That infrastructure can change its rules, change its algorithms, change its moderation systems, change its business priorities and change its interpretation of what content qualifies for monetisation. The artist who has built no alternative route to the audience is exposed every time the platform changes.

This is not an argument to abandon streaming.

It is an argument to stop depending on it.

Every independent artist should maintain their own master archive, production files, rights documentation, registrations, artwork, metadata and release history. Keep records that can demonstrate how the work was created. Preserve project files and stems where appropriate. Maintain documentation of collaborators and licences. Keep evidence of your rights. If a platform ever disputes your work, you should not be reconstructing your creative history from memory after the fact. You should already have the documentation.

And above all, build a direct route to your listeners.

A direct store changes the relationship because the artist controls the destination. The platform can remain useful for discovery, but the consumer has another way to support the creator. For UNIDARK, that route is the Official UNIDARK Store. The principle is simple: let streaming platforms compete for your presence rather than allowing your entire economic existence to depend upon their continued approval.

That distinction is the foundation of genuine independence.

If TIDAL wants artists to trust its AI policy, it should answer difficult questions publicly rather than relying primarily on terms and support forms. How accurate is its detector? How frequently are false positives observed? How many artists have successfully appealed AI classifications? How quickly are erroneous classifications corrected? How much royalty value has been withheld because of AI classification? How much has subsequently been restored after successful appeals? What evidence does TIDAL retain when making an AI determination? Does the artist receive meaningful information about the basis for the decision? How does TIDAL prevent a detector from confusing AI-assisted production with wholly AI-generated production?

These are not unreasonable demands.

They are basic accountability questions.

If TIDAL merely added an informational tag that had no financial consequence, the debate would still matter, but the stakes would be lower. Once the classification can determine whether a recording receives royalty attribution, the standard has to rise. Financial decisions require evidence. Financial decisions require correction mechanisms. Financial decisions require transparency. Financial decisions require records. Financial decisions require accountability.

TIDAL cannot reasonably demand that artists accept a potentially consequential classification while simultaneously telling them that the detection technology can produce false positives and that the company does not guarantee the accuracy of the label.

This is the philosophical problem underneath everything else. The independent musician should not have to pass an algorithmic test to establish that their creative labour is legitimate. If a platform believes that a recording is wholly AI-generated, it should have enough evidence to defend that conclusion when challenged. The burden should not simply fall onto the creator to demonstrate that the machine is wrong.

Otherwise, the platform has inverted the presumption.

The artist creates.

The machine suspects.

The corporation decides.

The artist appeals.

That is not artistic freedom.

That is algorithmic gatekeeping.

It is really about who owns the authority to define legitimate participation in the streaming economy. TIDAL has created rules around AI-generated music, artificial streams, impersonation and other forms of platform abuse. Some of those rules are unquestionably necessary. But every new rule also expands the corporation’s ability to decide what happens to content inside its ecosystem. That means artists should judge the platform not simply by its stated intentions, but by the safeguards surrounding its power.

The question should never be, “Does TIDAL mean well?”

The question should be, “What happens when TIDAL is wrong?”

That is the fact that should remain at the centre of this entire discussion. TIDAL’s own terms explicitly acknowledge the possibility of false positives and false negatives in AI detection. At the same time, TIDAL says that music it identifies as wholly AI-generated is not eligible for royalty attribution. Put those two statements together and the accountability problem becomes impossible to ignore: the company acknowledges that its detection system can make mistakes, while maintaining a policy in which its determination can affect whether music receives royalties.

That is not speculation.

That is the company’s own framework.

TIDAL has the power to scan. TIDAL has the power to classify. TIDAL has the power to label. TIDAL has the power to exclude content from royalty calculations. TIDAL has the power to remove certain content. TIDAL has the contractual protection of saying its detection is performed on a best-efforts basis. The artist has the responsibility to create the music, maintain the rights, distribute the recording, monitor the platform and challenge mistakes when they occur.

That is an extraordinary imbalance.

And independent artists should stop pretending it is normal.

TIDAL can be useful. Its audience can be useful. Its sound quality can be useful. Its distribution relationships can be useful. None of those benefits change the fundamental reality that TIDAL is a corporation operating a proprietary platform according to its own commercial rules. Its interests are not identical to yours. Its responsibility is to operate its service, manage its business, protect its systems and control its platform. Your responsibility is to protect your creative career.

Those responsibilities overlap.

They are not the same.

Do not wait until a platform labels your music incorrectly to discover that you have no direct audience. Do not wait until a distributor disputes a release to realise you have lost your production documentation. Do not wait until royalties disappear to start investigating how the accounting works. Do not wait until an algorithm changes to build a website. Build the infrastructure while everything is working.

Own what you can.

Document what you create.

Register what needs registration.

Archive everything.

Build direct relationships.

Sell directly.

Keep your audience informed.

And treat every streaming platform as a channel rather than a home.

If the company genuinely believes its AI policy protects artists, transparency should not threaten it. Publish meaningful accuracy information. Explain the appeal process. Explain the financial correction process. Explain what evidence is considered. Explain how false positives are handled. Explain how an artist’s lost royalties are restored when the company gets the classification wrong. Explain how human review works. Explain what “wholly” means operationally. Explain how the company distinguishes AI assistance from AI authorship.

If the system is good, transparency should strengthen it.

If the system cannot withstand scrutiny, artists have every reason to question it.

This distinction matters psychologically. Artists sometimes behave as though they are asking a powerful company for permission to participate in music. They are not. They are supplying creative work to a commercial ecosystem. The relationship should therefore be approached with the same caution any independent business owner applies when dealing with a large corporation. Read the terms. Understand the revenue chain. Understand who controls what. Maintain alternatives. Do not surrender your only route to customers.

TIDAL needs your music far more than your career needs TIDAL.

That is the mindset independent artists need.

A streaming profile can disappear from search results. A platform can change its interface. A policy can change. An algorithm can change. A corporation can change ownership. A detector can make a mistake. A distributor can make a mistake. A payment system can change. None of that changes the underlying value of your catalogue if you have preserved control over it.

Your masters are the asset.

Your compositions are the asset.

Your audience relationships are the asset.

Your documentation is the evidence.

The platform is the channel.

Never confuse the channel with the asset.

AI is not going away. Streaming is not going away. Automated moderation is not going away. Detection systems will become more sophisticated. Platforms will create more categories. Corporations will increasingly attempt to regulate the flood of synthetic content. Some of those developments will be necessary. Some will undoubtedly produce new problems.

The artists who survive will not necessarily be those who reject technology.

They will be those who refuse to become dependent on somebody else’s technology.

That is the lesson hidden inside the royalty dispute. The issue is not simply whether TIDAL’s detector works. The deeper issue is whether an artist’s economic survival should depend on a detector operated by a corporation that explicitly admits the technology can be wrong.

The answer should be obvious.

No independent artist should build their entire livelihood on a system they cannot inspect, cannot control and cannot guarantee will classify their work correctly.

Use TIDAL.

Use streaming.

Use discovery platforms.

But build something underneath them that belongs to you.

Because the moment a platform can tell you that your music is not eligible for royalties, and then tell you its technology might be wrong, the relationship has changed.

You are no longer simply distributing music.

You are negotiating with a gatekeeper.

And the strongest position an independent artist can have when dealing with a gatekeeper is the ability to walk away.

That is why direct ownership matters.

That is why direct sales matter.

That is why your own audience matters.

And that is why the safest place for your music will never be a platform you do not control.

UNIDARK, also known as Morning Star, is a UK-based independent extreme metal producer and the creator of Blackdeathgrin Metal — an original extreme metal genre combining elements of black metal, death metal, deathcore, and grindcore.

Explore the Official UNIDARK Hub

The central directory for music, releases, streaming platforms, videos, lyrics, Blackdeathgrin Metal information, and all official project links.
→ Official UNIDARK Hub

Explore Official Lyrics & Song Meanings

Discover the complete UNIDARK lyrics archive, song interpretations, and the concepts behind the music.
→ Lyrics Archive

Watch Official UNIDARK Music Videos

Experience the visual side of the project through official releases and video content.
→ Video Archive

Explore Metal Insights

Read more deep articles covering extreme metal genres, underground music, industry topics, production, platforms, culture, and the evolution of heavy music.
→ Metal Insights

Explore Sync Licensing

Read more deep articles covering music licensing, sync rights, music for film, television, games, advertising, trailers, independent production, hybrid music production, catalogue clearance, and the evolving relationship between music and visual media.
→ Sync Licensing

Explore the UNIDARK FAQ Archive

Find answers about UNIDARK, Morning Star, Blackdeathgrin Metal, releases, the creative process, and the philosophy behind the project.
→ UNIDARK FAQ

Support UNIDARK Directly

Support independent extreme metal creation through the official UNIDARK store.
→ Official UNIDARK Store

Join the Official UNIDARK Transmission

Subscribe for new releases, Metal Insights articles, official announcements and updates directly from UNIDARK. Follow the evolution of Blackdeathgrin Metal and the continuing development of the UNIDARK project.